Page 37 - Šolsko polje, XXVII, 2016, no. 3-4: IEA ICILS in druge sodobne teme, ur. Eva Klemenčič
P. 37
mirazchiyski ■ the digital divide ...
public (0.18) and Korea (0.18). The ones with the highest effects are Slo
venia (0.25), Croatia (0.26), Slovak Republic (0.26) and Poland (0.27).
However, the school SES on Level 2 has strong and significant effect in
all countries in ICILS 2013 and the coefficients are greater. The smallest
school-level SES effects are found in Korea (0.31) and Slovenia (0.37), the
largest are in Chile (0.82) and Poland (0.86).
To test for the compositional effect of school SES (difference betwe
en the school level effect and the person-level effect of SES), Model 2 was
further modified by centering SES around the grand mean on both Levels
1 and 2 (Model 3):
(L1, grand-mean centering) (4)
(L2, grand-mean centering)
(L2),
The results from these models are presented in Table 5. Statistical
ly significant compositional effects were found in 13 out of 20 countries:
Australia, Buenos Aires, Chile, Czech Republic, Denmark, Germany,
Hong Kong, Lithuania, Poland, Slovak Republic, Switzerland, Thailand
and Turkey. In all these countries the effect is strong and statistically sig
nificant, ranging from 0.43 (Turkey) to 0.73 (Chile). In the remaining sev
en countries (Canada [Newfoundland and Labrador], Canada [Ontario],
Croatia, Korea, Norway, Russian Federation and Slovenia) non-signifi
cant compositional effects was found.
Table 5. Model 3 standardized results for the compositional effect
Countries SES (L1) p (compositi- p
onal effect,
Australia 0.25 <0.001 <0.001
Buenos Aires (Argentina) 0.23 <0.001 L2) 0.005
Canada (Newfoundland and Labrador) 0.26 <0.001 0.57 0.079
Canada (Ontario) 0.22 <0.001 0.48 0.039
Chile 0.22 <0.001 0.29 <0.001
Croatia 0.30 <0.001 0.32 0.099
Czech Republic 0.21 <0.001 0.73 <0.001
Denmark† 0.25 <0.001 0.20 <0.001
Germany 0.07 0.295 0.60 <0.001
Hong Kong† -0.05 0.124 0.60 <0.001
0.72
0.45
35
public (0.18) and Korea (0.18). The ones with the highest effects are Slo
venia (0.25), Croatia (0.26), Slovak Republic (0.26) and Poland (0.27).
However, the school SES on Level 2 has strong and significant effect in
all countries in ICILS 2013 and the coefficients are greater. The smallest
school-level SES effects are found in Korea (0.31) and Slovenia (0.37), the
largest are in Chile (0.82) and Poland (0.86).
To test for the compositional effect of school SES (difference betwe
en the school level effect and the person-level effect of SES), Model 2 was
further modified by centering SES around the grand mean on both Levels
1 and 2 (Model 3):
(L1, grand-mean centering) (4)
(L2, grand-mean centering)
(L2),
The results from these models are presented in Table 5. Statistical
ly significant compositional effects were found in 13 out of 20 countries:
Australia, Buenos Aires, Chile, Czech Republic, Denmark, Germany,
Hong Kong, Lithuania, Poland, Slovak Republic, Switzerland, Thailand
and Turkey. In all these countries the effect is strong and statistically sig
nificant, ranging from 0.43 (Turkey) to 0.73 (Chile). In the remaining sev
en countries (Canada [Newfoundland and Labrador], Canada [Ontario],
Croatia, Korea, Norway, Russian Federation and Slovenia) non-signifi
cant compositional effects was found.
Table 5. Model 3 standardized results for the compositional effect
Countries SES (L1) p (compositi- p
onal effect,
Australia 0.25 <0.001 <0.001
Buenos Aires (Argentina) 0.23 <0.001 L2) 0.005
Canada (Newfoundland and Labrador) 0.26 <0.001 0.57 0.079
Canada (Ontario) 0.22 <0.001 0.48 0.039
Chile 0.22 <0.001 0.29 <0.001
Croatia 0.30 <0.001 0.32 0.099
Czech Republic 0.21 <0.001 0.73 <0.001
Denmark† 0.25 <0.001 0.20 <0.001
Germany 0.07 0.295 0.60 <0.001
Hong Kong† -0.05 0.124 0.60 <0.001
0.72
0.45
35
